@cncf.io"If it's shared, it's vulnerable."
At KubeCon + CloudNativeCon China, Zvonko Kaiser shared how Confidential Containers extend trust from silicon to service to protect AI workloads on Kubernetes.
How is your team handling confidential GPU workloads in production?
@cncf.ioBoosting dev capacity 15x on the same hardware sounds impossible until you rethink GPU scheduling.
CETC Cloud used Kubernetes + HAMi to scale from 2 to 30 concurrent dev Pods on an 8-GPU device with production AI.
@exoscale.comCloud-native conversations need a higher level. 😉
On Sep 28, SUSE and Exoscale host CloudCollective above Vienna, ahead of Cloud Native Days Austria 2026.
Kubernetes, platform engineering, digital sovereignty & AI-ready infrastructure.
Want a seat? Reach out!
@optimcloud.bsky.social☁️ OPTIM CLOUD: Sovereign infrastructure. AS207819 · 4.21M devices online · 99.998% uptime Talos K8s across AMS·FRA·LON·SIN·SYD·IAD OpenWrt, OpenWiFi, RDK-B, Plume — single pane of glass. https://www.optimcloud.com #optimcloud #kubernetes #talos #sovereign
@opsmatters.comThe latest update for #HAProxy includes "How to prove your HAProxy build is legitimate" and "#AI in the public sector (infrastructure challenges and solutions)".
#DevOps #Kubernetes #Security https://opsmtrs.com/3aGSzYy
@darryl-ruggles.cloud#Kubernetes #Storage #FinOps #PlatformEngineering
Kubernetes does not delete Persistent Volume Claims (PVCs) when their pods go away. This helps to protect the data but can leave orphaned volumes.
@0daybeats.bsky.socialWhat do sysadmins actually listen to while managing infra? From Proxmox clusters to K8s deploys, the definitive guide: https://www.0daybeats.com/blog/best-music-for-sysadmins #lofi #sysadmin #devops #cyberpunk #kubernetes
@devtocash.bsky.socialAlert on ScalingActive == false for 10m, not just latency. An HPA pinned at max with no metrics scaled up for a peak and is still billing you.
Full fix guide 👇
https://devtocash.com/blog/2026-09-22-kubernetes-hpa-unknown-targets-failedgetresourcemetric-fix
#Kubernetes #SRE
@hugovalters.bsky.socialCalico GlobalNetworkPolicy programs host eBPF/iptables via Felix to enforce a non-bypassable cluster network baseline across all namespaces.
#kubernetes #calico #ValtersIT
https://www.valtersit.com/vault/calico-global-network-policy-for-nonbypassable-cluster-basel-2d6f80/
@chrisshort.netKubernetes v1.37: Tracking When a PersistentVolumeClaim Was Last Used (Beta) #devopsish #kubernetes
@keyscope.devKeyscope v0.40.0 arrives tomorrow, introducing real-time resource watching and inspection for Kubernetes.
This initial release focuses on read-only visibility across essential workload, network, and configuration resources.
@akuity.ioArgo CD handles delivery. What about moving apps from dev → staging → production? 🔁
Join Kargo Maintainer Taylor Thomas for GitOps 101 and learn how Kargo brings automation and structure to promotions.
Register: buff.ly/O1kBjZP
@tellerstech.bsky.socialHelm 3 Reaches End of Life: What You Need to Know #helm #kubernetes #devops #sre #cloud This is a clip from our recent Ship It Weekly Podcast episode. Visit https://shipitweekly.fm or link in bio to listen to the full episode!
@opennebula.bsky.social🤝 Proud to be joining our close partners TrueFullstaq at #Edgecase2026 this week!
We’ll be bringing a keynote, two workshops, and our booth to Hilversum on September 24 🚀
Learn more about our collaboration: opennebula.io/case-studies...
@hexabase.bsky.socialWhy Managed K8s is the Answer in 2026: Understanding the 5 Stages of Infrastructure Evolution
@linuxnews.net[It's FOSS] Portainer Cuts the Cord Between Its Free and Paid Editions
@opennebula.bsky.social🚀 Automate Kubernetes on OpenNebula with Kubespray + Terraform.
Terraform provisions the infrastructure, while Kubespray deploys Kubernetes—with support for HA control planes, static IPs, and node anti-affinity.
👉 Read more: opennebula.io/blog/product...
@umbra-codex.bsky.socialA second cluster doesn't give you failover. Only a second place to run kubectl.
Karmada just graduated in the CNCF: placement, propagation and failover decided centrally, apps unchanged. A dying cluster becomes a scheduling decision, not a runbook.
@hugovalters.bsky.socialQuarantine a compromised pod fast: mutate labels to break Service selector matches, stop LB traffic, keep container alive for forensics, then apply NetworkPolicy. https://www.valtersit.com/vault/isolate-compromised-kubernetes-pod-with-network-policy-and-l-54ef6f/ #kubernetes #security #containment
@it-connect.bsky.social🌟 Portainer 3.0 : cette nouvelle version sera orientée pour Kubernetes, tandis que Docker et Podman sont relayés au second plan. La version CE quant à elle n'y aura pas le droit.
Plus d'infos ici :
- www.it-connect.fr/portainer-3-...
